Detailed studies of electron density changes in a mineral called langbeinite K₂Mg₂(SO₄)₃ under pressure have been performed. Single crystal X-ray data for this mineral under pressure (1GPa) were collected at the 13-BM-C beamline at the Advanced Photo Source (Argonne National Laboratory, USA). Additionally, complementary experiments at ambient conditions were performed on an in-house diffractometers. Experimental results were complemented by theoretical calculations within the pressure range up to 40 GPa.

From the point of view of mineralogical processes taking part in the Earth mantle (and the mantles of other even extraterrestrial planets), establishing detailed changes of electron density in minerals under pressure is absolutely crucial to understand the nature and mechanisms of mineralogical processes. Combining both experimental charge density studies and high pressure investigations is still a real challenge. This work is our continuation of our previous feasibility studies on experimental quantitative electron density investigations of electron density in grossular under 1GPa pressure [1].

Answering the questions how electron density distribution in langbeinite is affected by increasing pressure is obviously the main topic of this work. However there are also some other issues which we would like to address. Are there any differences between experimental and theoretical charge density distributions obtained on the basis of experimental data and theoretical dynamic structure factors? Are there any significant differences in properties of charge density distributions obtained for complete and incomplete high resolution X-ray diffraction data sets? Are there any differences in charge density distributions obtained for X-ray data collected with two different wavelengths of X-ray radiation? Should the data be absolutely complete to obtain reasonable experimental charge density distribution? When experimental data are impossible to be collected, is it reasonable to use theoretically calculated dynamic structure factors instead and refine theoretical models of electron density?

Langbeinite crystalizes in the cubic P2₁3 space group. Its structure is composed of SO₄ tetrahedra and MgO₆ octahedra. Potassium cations which are placed in the voids between these polyhedra are surrounded by oxygen anions. Unfortunately due to significant deformation, one cannot say that KO₁₂ is a regular icosahedron. Although mentioned polyhedra seem to completely fill in the space, this schematic way of presentation is not the best one when topology of electron density distribution must be described.

Investigating changes of electron density as a function of pressure, we are going to compare electron density properties at BCPs, integrated atomic basins, changes of thermal ellipsoids. Obviously, raising pressure will cause shrinking of the unit cell and consequently, changes of electron density distribution. However, the question is how exactly such changes will manifest.

No doubt that polyhedra commonly used in mineralogy and crystallography are not useful representation of electron density as they neither have full representation of electron density of the central ion nor any of the corners ions. So from time to time returns an old question: how big are atoms in crystals [2]. Here we will answer this question and the other ones already mentioned above at the level of quantitative electron density distributions in our model mineral.

[1] Gajda, R., Stachowicz, M., Makal, A., Sutuła, S., Parafiniuk, J., Fertey, P. & Woźniak, K. (2020). IUCrJ. 7, 383-392.

[2] Brown, I. D. (2017). Struct. Chem. 28, 1377-1387.

Keywords: high pressure; electron density; theoretical structure factors

A quantitative experimental charge density study was undertaken for the double antiperovskite mineral – sulphohalite [Na₆(SO₄)₂FCl]. High-resolution X-ray diffraction data was collected employing AgKα radiation (λ = 0.56087 Å) to a resolution of 0.3941 Å at 100K. Electron density (ED) distribution – ρ(r) was modelled, in compliance with the Hansen-Coppens formalism[1], by consecutive least-square multipolar refinements. Based on such experimental distribution of charge, QTAIM topological analysis[2] was undertaken. Full-volume property integration over delineated atomic basins (AB’s) yielded their appertaining charges [Q_AB-Cl = -0.836e^-; Q_AB-S = 03.168e^-; Q_AB-Na = 0.910e^-; Q_AB-F = -1.334e^-; and Q_AB-O = -1.227e^-] and volumes [V_AB-Cl = 38.920ų; V_AB-S = 5.656ų; V_AB-Na = 7.931ų; V_AB-F = 14.178 ų and V_AB-O = 17.416 ų]. The percentage of unaccounted electrons and volume per unit cell was respectively 0.010% and 0.406%. Within the uncertainty range of performed numerical integration, such percentages can be unheeded. A total of 6·BCP’s [∇²ρ(r_Cl···S) = 0.120e^-·Å^-5; ∇²ρ(r_Cl···Na) = 0.575e^-·Å^-5; ∇²ρ(r_S-O) = -31.00e^-·Å^-5; ∇²ρ(r_Na···O) = 1.931e^- ·Å^-5; ∇²ρ(r_Na···F) = 3.022e^-·Å^-5 and ∇²ρ(r_F···O) = 0.868e^-·Å^-5], 5·RCP’s [∇²ρ(r_I) = 0.912e^-·Å^-5; ∇²ρ(r_II) = 0.332e^-·Å^-5 and ∇²ρ(r_III,IV,V) = 0.201e^-·Å^-5] and 4·CCP’s [∇²ρ(r_I,II) = 0.514e^-·Å^-5 and ∇²ρ(r_III,IV) = 0.401e^-·Å^-5] were identified (Figure 1). Hence, Morse’s ‘characteristic set’ condition was met[3]. The study of primary bundles (PB’s), as proposed by Pendás[4], revealed the interconnection between AB’s and CP’s onto basins of attraction or basins of repulsion. The nature of interatomic interactions was assessed through the dichotomous classification[3]. The S–O contact was acknowledged as a covalent with a shared-shell. The remaining contacts were characterized as non-covalent closed-shell (Cl···Na, Na···O and Na···F) or weak van der Waals closed-shell (Cl···S and F···O).

Modern approaches of X-ray diffraction allow for detailed quantitative studies of electron density in crystals of minerals. They can be combined with high pressure studies [1] as we demonstrate in this work for model zeolite mineral hsianqhualite, Ca₃Li₂(Be₃Si₃O₁₂)F₂.

At the level of electron density analysis first order configurational components in crystal structure description (Fig. 1a) were replaced by Bader’s atomic basins [2] which quantitatively characterise electron density of particular ions in mineral structures as well as precisely defined space, they occupy (Fig 1b). Their anisotropic and highly non-spherical shape reflects interatomic interactions and is sensitive to applied pressure. According to our studies the charge of ions in the crystal lattice differ from the formal, integer values and when external pressure is applied a redistribution of charge among ions takes place. This redistribution changes the size and shape, mostly at the edges of ionic basins in nonbonding fragments (Fig. 1c, d).

Negative compressibility of the F ion was observed. It was caused by the flow of electrons increasing the total negative charge and, consequently, increasing the volume of F ion at 1.9 GPa pressure (Fig 1d). Also inside of atomic basins of atoms electron density redistributes notably due to pressure.

The quantitative characterization of minerals under high pressure at the subatomic level of electron density, rise possibilities to better understand the nature of mineralogical process, phase transitions and formation of new phases and also to study plastic deformations of minerals using diamond anvil cells.

Zeolites are commonly used as ion-exchange materials for the remediation of nuclear waste; however, they have certain drawbacks. Unlike zeolites which contain SiO₄ and AlO₄ tetrahedra, microporous Ti-silicates can contain SiO₄ tetrahedra and TiO₆ octahedra and therefore structures are possible which have no traditional aluminosilicate analogues ^[1]. Microporous Ti-silicates such as sitinakite KNa₂Ti₄Si₂O₁₃(OH)·4H₂O and the synthetic niobium doped analogue are used for the removal of Cs⁺ and Sr²⁺ from nuclear waste ^[2,3]. The work presented here will focus on the structures and thermal behaviour of the ion-exchanged Ti- and Zr-silicates. A clear understanding of both is fundamental in determining if these materials have potential as ion-exchangers within the nuclear industry.

Umbite is a naturally occurring small pore microporous Zr- silicate, found in northern Russia and synthetic analogues, K₂ZrSi₃O₉·H₂O, can be prepared in the laboratory ^[4]. Ion-exchange studies here have shown that umbite has a preference for common radionuclides, such as Cs⁺ and Sr²⁺and Ce⁴⁺ (as a surrogate for Pu), even in the presence of competing ions. In-situ studies show that these materials behave differently with temperature, indicating that the nature and location of the charge balancing cation plays an important part in determining which high temperature phases are formed and the phases formed do not fit previously reported structures.

Natisite is another material which has interesting ion-exchange chemistry and is a layered Ti-silicate with the formula Na₂TiSiO₅ ^[6]_. The structure consists of square pyramidal titanium, with the sodium cations located between the layers. This coordination environment is highly unusual for Ti. Inclusion of zirconium or vanadium in the framework has a considerable effect on the ion-exchange properties, with changes in the exchange capacity and the rate of uptake for certain ions of interest.

A combination of techniques to probe long and short-range order (PDF and XAS) have been used to understand the ion-exchange and thermal behaviour of these materials.

References:

1) P. A. Wright, Microporous Framework Solids, The Royal Society of Chemistry, Cambridge, 2008. 2) D. M. Poojary, et al., Chem. Mater., 6, 2364 (1994). 3) A. Tripathi, et al., J. Solid State Chem., 175, 72 (2003). 4) D. M. Poojary, et al., Inorg. Chem., 36, 3072 (1997). 5) A. Ferreira, et al., J. Solid State Chem., 183, 3067 (2010). 6) D.G. Medvedev et al., Chem. Mater., 16, 3659 (2004).

The synthesis of multifunctional materials is a hot focus of research in materials science. In this respect, the synthesis of complexes based on the combination of organic-inorganic building blocks provides a promising approach in the design of systems with tuneable properties. In this communication we will present the properties of a new compound based on quinuclidine as the organic cation and FeCl₄ as the inorganic anion, with the formula (quinuclidine)[FeCl₄]. Similar compounds derived of this heterocyclic cation have been found to present interesting ferroelectric properties.[1] In this context, the multifunctional behaviour of this novel molecular crystal is related to the electronic structure of the 3d⁵ configuration of the Fe(III) ions together with the ability of the counter-ions to change of orientation or even become disordered as a function of temperature.

The structural characterization of (quinuclidine)[FeCl₄] compound shows two phase transitions. The first one, detected in the range from 100 to 300 K, was resolved by single-crystal X-Ray and neutron diffraction. At 300 K, the compound presents the orthorhombic space group Pbc2₁. At 100 K, the space group is Pbca, with a doubling of the a-axis, related to the rotation of the cations: two different orientations of the counterion are observed in the low temperature phase along the a direction, contrary to the high temperature phase, where it appears only one orientation.

Moreover, this compound presents long-range magnetic order below 3 K. The magnetic structure was solved using single-crystal and powder neutron diffraction data from D19 and D1B instruments (ILL, France), respectively. Our best model was found on the Shubnikov magnetic space group P2₁’2₁’2_1. Although the refined model present an antiferromagnetic structure, based on the symmetry analysis of the P2₁’2₁’2₁ Shubnikov group, a ferromagnetic component along the c direction is allowed. However, the refinement of this ferromagnetic component is beyond the precision of our measurements. Nevertheless, this can be fixed to the values derived from the macroscopic magnetometry measurements (SQUID). In order to provide a complete model these values were included in the magnetic model and fixed during the refinements.

At temperatures higher than R.T, there is a second structural phase transition which produces an important modification of the electric behaviour, as it has been reported on similar compounds.[1] The dielectric permittivity data collected shows a sharp phase transition around 390 K (also observed in DSC measurements). The value of the permittivity increases drastically with the increase of the temperature, reaching a maxima of 10⁵ at 390 K (measured at 1 kHz). This value is notable larger than similar compounds of this family.[1] This interesting behaviour could be of interest for electrochemical applications.

[1] (a) Jun Harada et al., Nat Chem, 2016 Oct; 8(10):946-52. (b) You-Meng You et al., Nat Commun. 2017, 8:14934. c) Ting Fang et al., Z. Anorg. Allg. Chem. 2019, 645, 3–7. d) Guang-Meng Fan et al., CrystEngComm, 2018,20, 7058-7061.